Work Package 3

Tuesday 5 June 2012
by  daubin
popularity : 0%

Coordinator: Hugues Roest-Crollius (CNRS, ENS Paris)

Contribution: LBBE


Ancestral genome reconstructions, at the level of the nucleotidic sequence, gene order or karyotype, have been achieved in numerous taxonomic groups including yeasts, plants, fishes and mammals. They potentially represent unique points of reference from which the study of the evolution of molecular properties along descending branches becomes possible. However, no generic method has yet been established that will manage the large volume of genomic data being generated by the community (>100 metazoans, >20 plants, > 1000 prokaryotes). Partners 3 and 1 have already developed methods to reconstruct the ancestral order of genes (Muffato et al. 2010; Chauve et al. 2010; Murat et al. 2010), which provide a strong basis for the proposed work.
We plan to build on the models developed in WP2 to adopt a probabilistic approach that will add key features to the former methods:

  1. it will provide statistical support for the inference of ancestral linkages, both globally at the level of chromosome and genome reconstructions and locally at the level of gene-to-gene adjacencies;
  2. it will exploit a tight coupling between gene phylogeny reconstruction and gene order reconstruction and will introduce dependencies between chronologically successive ancestors. Indeed current methods target a single ancestor at a time, and are thus blind to potential inconsistencies between successive ancestors;
  3. probabilistic models will allow an inclusion in the context given by WP1, and thus a feedback for the improvement of gene phylogenies.
  4. the model will be sufficiently generic to reconstruct the order of any biological entity that possesses chromosome coordinates: protein coding genes, RNA gene, pseudogene, transposable element, regulatory region, etc.

The method will be applied to 100’s of sequenced genomes in order to provide a uniform resource comparable across taxa, with known confidence measures and at a resolution that allows meaningful biological inferences. Those will concern the functional role of the chromosome organization, which is a highly debated scientific topic. Beyond the operon structure of most Bacteria and some eukaryotes (Lathe, Snel, et Bork 2000; Boccard, Esnault, et Valens 2005; Bailly-Bechet et al. 2006) the comparison of the ordered sequences of genes from ancestors to descendants would allow an estimation of how large-scale structures evolve. In turn, such observations should allow us to make conclusions on hypotheses concerning their function. In particular the correlation between this organization and the other forms of complex molecular systems inferred from WP4-6 will allow us to decipher the respective roles of historical contingencies and functional constrains in shaping the extant chromosomes.
LBBE (E. Tannier) and ENS (H. Roest-Crollius) are specifically qualified to form a team around this subject, both having developed methods for ancestral genome reconstructions and applied them to yeast (Chauve et al. 2010), plants (Murat et al. 2010), mammals, or a wide range of eukaryotes (Muffato et al. 2010).